COVID vs Despair

This is an R Markdown Notebook. When you execute code within the notebook, the results appear beneath the code.

Try executing this chunk by clicking the Run button within the chunk or by placing your cursor inside it and pressing Ctrl+Shift+Enter.

library(ggplot2)
library(dplyr)

cdc.data

despair.data<- subset(cdc.data, death_cause == "Despair")

#covid_data_states<- read.csv("covid_data_states.csv")

covid_TS_counties_long.cases

keeps.despair <- c("county_name","county_fips","death_rate")
despair.data.updated = despair.data[keeps.despair]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated = covid_TS_counties_long.cases[keeps.covid]

COVIDdespair.all = merge(despair.data.updated, covid.data.updated, by.x="county_fips", by.y="countyFIPS")

COVIDdespair<- distinct(COVIDdespair.all)

ggplot(data=COVIDdespair, aes(x=death_rate,y=p_deaths)) +
  geom_point()

cor.test(COVIDdespair$death_rate,COVIDdespair$p_deaths)

ggplot(data=COVIDdespair, aes(x=death_rate, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE)

library(gridExtra)
#>50
cdc.data

#despair.data<- subset(cdc.data, death_cause == "Despair", period == "2015-2017")
despair.data <- cdc.data[ which(cdc.data$death_cause=='Despair'
& cdc.data$period== '2015-2017'), ]

covid_TS_counties_long.cases

#coviddeathallcounties1<- subset(covid_TS_counties_long.cases, date == "2020-03-15")
#coviddeathallcounties <- subset(coviddeathallcounties1, deaths >= 5, select=c("County","countyFIPS","deaths", "p_deaths"))
coviddeathallcounties <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 50 ),]
#& covid_TS_counties_long.cases$date=="2020-03-15"), ]
#coviddeathallcounties[coviddeathallcounties$date >= "2020-04-01" & coviddeathallcounties$date <= "2020-04-03",]

keeps.despair <- c("county_name","county_fips","death_rate")
despair.data.updated = despair.data[keeps.despair]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated = coviddeathallcounties[keeps.covid]

COVIDdespair.all = merge(despair.data.updated, covid.data.updated, by.x="county_fips", by.y="countyFIPS")


#COVIDdespair<- distinct(COVIDdespair.all)
COVIDdespair<- distinct(COVIDdespair.all,county_fips, .keep_all= TRUE)

#COVIDdespair <- unique(COVIDdespair[ , 1:3 ] )

#ggplot(data=COVIDdespair, aes(x=death_rate,y=p_deaths)) +
#  geom_point()

cor.test(COVIDdespair$death_rate,COVIDdespair$p_deaths)

    Pearson's product-moment correlation

data:  COVIDdespair$death_rate and COVIDdespair$p_deaths
t = -0.90027, df = 311, p-value = 0.3687
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.16093448  0.06021813
sample estimates:
        cor 
-0.05098315 

#-0.05098315 

ggplot(data=COVIDdespair, aes(x=death_rate, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (>50)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

#0-50
cdc.data

despair.data <- cdc.data[ which(cdc.data$death_cause=='Despair'
& cdc.data$period== '2015-2017'), ]

covid_TS_counties_long.cases

coviddeathallcountiesunder50 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths <=50 ),]

keeps.despairunder50 <- c("county_name","county_fips","death_rate")
despair.data.updated.under50 = despair.data[keeps.despairunder50]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated = coviddeathallcountiesunder50[keeps.covid]

COVIDdespair.all.under50 = merge(despair.data.updated.under50, covid.data.updated, by.x="county_fips", by.y="countyFIPS")

COVIDdespairunder50<- distinct(COVIDdespair.all.under50,county_fips, .keep_all= TRUE)

cor.test(COVIDdespairunder50$death_rate,COVIDdespairunder50$p_deaths)

    Pearson's product-moment correlation

data:  COVIDdespairunder50$death_rate and COVIDdespairunder50$p_deaths
t = -4.0265, df = 2703, p-value = 5.818e-05
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.11456919 -0.03964255
sample estimates:
        cor 
-0.07721489 

#-0.07721489 
p1<-ggplot(data=COVIDdespairunder50, aes(x=death_rate, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (0-50)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

log10.covid.0_50<-(log10(COVIDdespairunder50$p_deaths))
log10.despair.0_50<- (log10(COVIDdespairunder50$death_rate))
COVIDdespairunder50$log10covid= log10.covid.0_50
COVIDdespairunder50$log10despair=log10.despair.0_50
p1_5<-ggplot(data=COVIDdespairunder50, aes(x=log10despair, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (0-50)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

#50-100
cdc.data

despair.data <- cdc.data[ which(cdc.data$death_cause=='Despair'
& cdc.data$period== '2015-2017'), ]

covid_TS_counties_long.cases

coviddeathallcounties50100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >=50|covid_TS_counties_long.cases$deaths <=100 ),]

keeps.despair50100 <- c("county_name","county_fips","death_rate")
despair.data.updated.50100 = despair.data[keeps.despair50100]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated50100 = coviddeathallcounties50100[keeps.covid]

COVIDdespair.all.50100 = merge(despair.data.updated.50100, covid.data.updated50100, by.x="county_fips", by.y="countyFIPS")

COVIDdespair50100<- distinct(COVIDdespair.all.50100,county_fips, .keep_all= TRUE)

cor.test(COVIDdespair50100$death_rate,COVIDdespair50100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDdespair50100$death_rate and COVIDdespair50100$p_deaths
t = -1.3486, df = 2703, p-value = 0.1776
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.06355648  0.01176832
sample estimates:
        cor 
-0.02593088 

#-0.02593088 
p2<-ggplot(data=COVIDdespair50100, aes(x=death_rate, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (50-100)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

log10.covid.50_100<-(log10(COVIDdespair50100$p_deaths))
log10.despair.50_100<- (log10(COVIDdespair50100$death_rate))
COVIDdespair50100$log10covid= log10.covid.50_100
COVIDdespair50100$log10despair=log10.despair.50_100
p2_5<-ggplot(data=COVIDdespair50100, aes(x=log10despair, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (50-100)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

#>100
cdc.data

despair.data <- cdc.data[ which(cdc.data$death_cause=='Despair'
& cdc.data$period== '2015-2017'), ]

covid_TS_counties_long.cases

coviddeathallcountiesgreater100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >=100 ),]

keeps.despairgreater100 <- c("county_name","county_fips","death_rate")
despair.data.updated.greater100 = despair.data[keeps.despairgreater100]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updatedgreater100 = coviddeathallcountiesgreater100[keeps.covid]

COVIDdespair.all.greater100 = merge(despair.data.updated.greater100, covid.data.updatedgreater100, by.x="county_fips", by.y="countyFIPS")

COVIDdespairgreater100<- distinct(COVIDdespair.all.greater100,county_fips, .keep_all= TRUE)

cor.test(COVIDdespairgreater100$death_rate,COVIDdespairgreater100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDdespairgreater100$death_rate and COVIDdespairgreater100$p_deaths
t = -0.97918, df = 187, p-value = 0.3288
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.21199107  0.07204285
sample estimates:
        cor 
-0.07142184 

#-0.07142184
p3<-ggplot(data=COVIDdespairgreater100, aes(x=death_rate, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (>100)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

log10.covid.greater100<-(log10(COVIDdespairgreater100$p_deaths))
log10.despair.greater100<- (log10(COVIDdespairgreater100$death_rate))
COVIDdespairgreater100$log10covid= log10.covid.greater100
COVIDdespairgreater100$log10despair=log10.despair.greater100
p3_5<-ggplot(data=COVIDdespairgreater100, aes(x=log10despair, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Deaths of Despair Rate vs COVID Mortality Rate (>100)") +
  xlab("Death of Despair Rate") +
  ylab("COVID Mortality Rate")

grid.arrange(p1, p2, p3, nrow = 3)

grid.arrange(p1_5,p2_5,p3_5, nrow =3)

p1

p2

p3

p1_5

p2_5

p3_5

cor.test(COVIDdespairunder50$log10covid,COVIDdespairunder50$log10despair)
cor.test(COVIDdespair50100$log10covid,COVIDdespair50100$log10despair)
cor.test(COVIDdespairgreater100$log10covid,COVIDdespairgreater100$log10despair)

suicide

library(tidyverse)
library(dplyr)

aaa<-read.csv("2020CHR.csv")

data.suicide <- aaa %>% dplyr::select(FIPS, State, County, Suicide.Rate..Age.Adjusted.)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcounties <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 50 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated = coviddeathallcounties[keeps.covid]

COVIDsuicide.all = merge(data.suicide, covid.data.updated, by.x="FIPS", by.y="countyFIPS")

COVIDsuicide<- distinct(COVIDsuicide.all,FIPS, .keep_all= TRUE)

cor.test(COVIDsuicide$Suicide.Rate..Age.Adjusted.,COVIDsuicide$p_deaths)

    Pearson's product-moment correlation

data:  COVIDsuicide$Suicide.Rate..Age.Adjusted. and COVIDsuicide$p_deaths
t = -2.7007, df = 368, p-value = 0.007239
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.23798074 -0.03799566
sample estimates:
       cor 
-0.1394094 

#-0.139

ggplot(data=COVIDsuicide, aes(x=Suicide.Rate..Age.Adjusted., y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (>50)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

#50
aaa<-read.csv("2020CHR.csv")

data.suicide <- aaa %>% dplyr::select(FIPS, State, County, Suicide.Rate..Age.Adjusted.)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcountiesunder50 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths <= 50 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updatedunder50 = coviddeathallcountiesunder50[keeps.covid]

COVIDsuicide.all.under50 = merge(data.suicide, covid.data.updatedunder50, by.x="FIPS", by.y="countyFIPS")

COVIDsuicide.under50<- distinct(COVIDsuicide.all.under50,FIPS, .keep_all= TRUE)

cor.test(COVIDsuicide.under50$Suicide.Rate..Age.Adjusted.,COVIDsuicide.under50$p_deaths)

    Pearson's product-moment correlation

data:  COVIDsuicide.under50$Suicide.Rate..Age.Adjusted. and COVIDsuicide.under50$p_deaths
t = -5.1303, df = 2367, p-value = 3.127e-07
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.14453012 -0.06486999
sample estimates:
       cor 
-0.1048683 

#-0.139

g1<-ggplot(data=COVIDsuicide.under50, aes(x=Suicide.Rate..Age.Adjusted., y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (0-50)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

log10.covid.under50<-(log10(COVIDsuicide.under50$p_deaths))
COVIDsuicide.under50$log10covid= log10.covid.under50
g1_5<-ggplot(data=COVIDsuicide.under50, aes(x=Suicide.Rate..Age.Adjusted., y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (0-50)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

#50-100
aaa<-read.csv("2020CHR.csv")

data.suicide <- aaa %>% dplyr::select(FIPS, State, County, Suicide.Rate..Age.Adjusted.)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcounties50_100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 50|covid_TS_counties_long.cases$deaths <= 100 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated50_100 = coviddeathallcounties50_100[keeps.covid]

COVIDsuicide.all.50_100 = merge(data.suicide, covid.data.updated50_100, by.x="FIPS", by.y="countyFIPS")

COVIDsuicide.50_100<- distinct(COVIDsuicide.all.50_100,FIPS, .keep_all= TRUE)

cor.test(COVIDsuicide.50_100$Suicide.Rate..Age.Adjusted.,COVIDsuicide.50_100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDsuicide.50_100$Suicide.Rate..Age.Adjusted. and COVIDsuicide.50_100$p_deaths
t = -8.149, df = 2367, p-value = 5.876e-16
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.2041096 -0.1257597
sample estimates:
       cor 
-0.1651953 

#-0.139

g2<-ggplot(data=COVIDsuicide.50_100, aes(x=Suicide.Rate..Age.Adjusted., y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (50-100)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

log10.covid.50_100<-(log10(COVIDsuicide.50_100$p_deaths))
COVIDsuicide.50_100$log10covid= log10.covid.50_100
g2_5<-ggplot(data=COVIDsuicide.50_100, aes(x=Suicide.Rate..Age.Adjusted., y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (50-100)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

#>100
aaa<-read.csv("2020CHR.csv")

data.suicide <- aaa %>% dplyr::select(FIPS, State, County, Suicide.Rate..Age.Adjusted.)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcountiesgreater100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 100 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updatedgreater100 = coviddeathallcountiesgreater100[keeps.covid]

COVIDsuicide.all.greater100 = merge(data.suicide, covid.data.updatedgreater100, by.x="FIPS", by.y="countyFIPS")

COVIDsuicide.greater100<- distinct(COVIDsuicide.all.greater100,FIPS, .keep_all= TRUE)

cor.test(COVIDsuicide.greater100$Suicide.Rate..Age.Adjusted.,COVIDsuicide.greater100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDsuicide.greater100$Suicide.Rate..Age.Adjusted. and COVIDsuicide.greater100$p_deaths
t = -2.114, df = 227, p-value = 0.03561
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.263839278 -0.009480459
sample estimates:
       cor 
-0.1389508 

#-0.139

g3<-ggplot(data=COVIDsuicide.greater100, aes(x=Suicide.Rate..Age.Adjusted., y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (>100)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

log10.covid.greater100<-(log10(COVIDsuicide.greater100$p_deaths))
COVIDsuicide.greater100$log10covid= log10.covid.greater100
g3_5<-ggplot(data=COVIDsuicide.greater100, aes(x=Suicide.Rate..Age.Adjusted., y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("Suicide Rate (Age Adjusted) vs COVID Mortality Rate (>100)") +
  xlab("Suicide Rate (Age Adjusted)") +
  ylab("COVID Mortality Rate")

g1

g2

g3

g1_5

g2_5

g3_5

grid.arrange(g1, g2, g3, nrow = 3)

grid.arrange(g1_5,g2_5,g3_5, nrow=3)

cor.test(COVIDsuicide.under50$Suicide.Rate..Age.Adjusted.,COVIDsuicide.under50$log10covid)
cor.test(COVIDsuicide.50_100$Suicide.Rate..Age.Adjusted.,COVIDsuicide.50_100$log10covid)
cor.test(COVIDsuicide.greater100$Suicide.Rate..Age.Adjusted.,COVIDsuicide.greater100$log10covid)

excessive drinking

library(tidyverse)
library(dplyr)
library(ggplot2)

#0-50
aaa<-read.csv("2020CHR.csv")

data.excessdrink <- aaa %>% dplyr::select(FIPS, State, County,X..Excessive.Drinking)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcounties <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths <= 50 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated = coviddeathallcounties[keeps.covid]

COVIDexcessdrink.all = merge(data.excessdrink, covid.data.updated, by.x="FIPS", by.y="countyFIPS")

COVIDexcessdrink<- distinct(COVIDexcessdrink.all,FIPS, .keep_all= TRUE)

cor.test(COVIDexcessdrink$X..Excessive.Drinking,COVIDexcessdrink$p_deaths)

    Pearson's product-moment correlation

data:  COVIDexcessdrink$X..Excessive.Drinking and COVIDexcessdrink$p_deaths
t = -8.3311, df = 3004, p-value < 2.2e-16
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.1850335 -0.1151447
sample estimates:
       cor 
-0.1502768 

gg1<-ggplot(data=COVIDexcessdrink, aes(x=X..Excessive.Drinking, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (<50)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

log10.covid.greater100<-(log10(COVIDexcessdrink$p_deaths))
COVIDexcessdrink$log10covid= log10.covid.greater100
gg1_5<-ggplot(data=COVIDexcessdrink, aes(x=X..Excessive.Drinking, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (<50)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

#50-100
aaa<-read.csv("2020CHR.csv")

data.excessdrink <- aaa %>% dplyr::select(FIPS, State, County, X..Excessive.Drinking)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcounties50_100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 50|covid_TS_counties_long.cases$deaths <= 100 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updated50_100 = coviddeathallcounties50_100[keeps.covid]

COVIDexcessdrink.all.50_100 = merge(data.excessdrink, covid.data.updated50_100, by.x="FIPS", by.y="countyFIPS")

COVIDexcessdrink.50_100<- distinct(COVIDexcessdrink.all.50_100,FIPS, .keep_all= TRUE)

cor.test(COVIDexcessdrink.50_100$X..Excessive.Drinking,COVIDexcessdrink.50_100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDexcessdrink.50_100$X..Excessive.Drinking and COVIDexcessdrink.50_100$p_deaths
t = -5.3865, df = 3004, p-value = 7.738e-08
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.13309280 -0.06227442
sample estimates:
        cor 
-0.09780742 

gg2<-ggplot(data=COVIDexcessdrink.50_100, aes(x=X..Excessive.Drinking, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (50-100)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

log10.covid.50_100<-(log10(COVIDexcessdrink.50_100$p_deaths))
COVIDexcessdrink.50_100$log10covid= log10.covid.50_100
gg2_5<-ggplot(data=COVIDexcessdrink.50_100, aes(x=X..Excessive.Drinking, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (50-100)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

#>100
aaa<-read.csv("2020CHR.csv")

data.excessdrink <- aaa %>% dplyr::select(FIPS, State, County, X..Excessive.Drinking)

covid_TS_counties_long.cases<- read.csv("covid_TS_counties_long.cases.csv")

coviddeathallcountiesgreater100 <- covid_TS_counties_long.cases[ which(covid_TS_counties_long.cases$deaths >= 100 ),]

keeps.covid <- c("County","countyFIPS","p_deaths")
covid.data.updatedgreater100 = coviddeathallcountiesgreater100[keeps.covid]

COVIDexcessdrink.all.greater100 = merge(data.excessdrink, covid.data.updatedgreater100, by.x="FIPS", by.y="countyFIPS")

COVIDexcessdrink.greater100<- distinct(COVIDexcessdrink.all.greater100,FIPS, .keep_all= TRUE)

cor.test(COVIDexcessdrink.greater100$X..Excessive.Drinking,COVIDexcessdrink.greater100$p_deaths)

    Pearson's product-moment correlation

data:  COVIDexcessdrink.greater100$X..Excessive.Drinking and COVIDexcessdrink.greater100$p_deaths
t = -1.3232, df = 227, p-value = 0.1871
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
 -0.21469091  0.04264009
sample estimates:
        cor 
-0.08748468 

gg3<-ggplot(data=COVIDexcessdrink.greater100, aes(x=X..Excessive.Drinking, y=p_deaths)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (>100)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

log10.covid.greater100<-(log10(COVIDexcessdrink.greater100$p_deaths))
COVIDexcessdrink.greater100$log10covid= log10.covid.greater100
gg3_5<-ggplot(data=COVIDexcessdrink.greater100, aes(x=X..Excessive.Drinking, y=log10covid)) + 
  geom_point() + geom_smooth(method="lm",se= TRUE) +
  ggtitle("% Excessive Drinking vs COVID Mortality Rate (>100)") +
  xlab("% Excessive Drinking") +
  ylab("COVID Mortality Rate")

gg1

gg2

gg3

gg1_5

gg2_5

gg3_5

grid.arrange(gg1, gg2, gg3, nrow = 3)

grid.arrange(gg1_5,gg2_5,gg3_5, nrow=3)

cor.test(COVIDexcessdrink$X..Excessive.Drinking,COVIDexcessdrink$log10covid)
cor.test(COVIDexcessdrink.50_100$X..Excessive.Drinking,COVIDexcessdrink.50_100$log10covid)
cor.test(COVIDexcessdrink.greater100$X..Excessive.Drinking,COVIDexcessdrink.greater100$log10covid)